[llvm] [AMDGPU] Generate waterfall for calls with SGPR(inreg) argument (PR #146997)

[Matt Arsenault via llvm-commits]

================
@@ -1,22 +1,452 @@
-; RUN: not llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs=0 -filetype=null %s 2>&1 | FileCheck -enable-var-scope %s
-
-; CHECK: illegal VGPR to SGPR copy
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -enable-var-scope %s
 
 declare hidden void @external_void_func_a15i32_inreg([15 x i32] inreg) #0
 declare hidden void @external_void_func_a16i32_inreg([16 x i32] inreg) #0
 declare hidden void @external_void_func_a15i32_inreg_i32_inreg([15 x i32] inreg, i32 inreg) #0
 
 define void @test_call_external_void_func_a15i32_inreg([15 x i32] inreg %arg0) #0 {
+; CHECK-LABEL: test_call_external_void_func_a15i32_inreg:
+; CHECK:       ; %bb.0:
+; CHECK-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; CHECK-NEXT:    s_mov_b32 s40, s33
+; CHECK-NEXT:    s_mov_b32 s33, s32
+; CHECK-NEXT:    s_or_saveexec_b64 s[42:43], -1
+; CHECK-NEXT:    buffer_store_dword v40, off, s[0:3], s33 ; 4-byte Folded Spill
+; CHECK-NEXT:    s_mov_b64 exec, s[42:43]
+; CHECK-NEXT:    v_writelane_b32 v40, s40, 32
----------------
arsenm wrote:

If you added some trivial bodies for the functions, IPRA would help trim out a lot of these CSR spills and reduce the test output to more relevant cases 

https://github.com/llvm/llvm-project/pull/146997